Aria Sinsonido preamp repair

[ef37a]

Now that's an expression I've not heard before - must be one of those Northern things

Anyway, I'm quite happy with the idea.

Oh! Yes, getting "bolted up" is being volunteered/pressed into a task that you did not want and is not actually in your job descrip'!

Not to be confused with "fitted up" which is something our police never do or ever did.

Dave.

[Dr R]

If you really are stuck, send me a priv message - I'm a sucker for an unusual challenge

In that case I most definitely will, thank you.

Rich

[Folderol]

This is proving to be a very strange design indeed! I've worked out a drawing, and it's so strange I had to keep re-checking it
See here.

(Updated diagram shown above)
Cold checks with a meter suggest there are no faults, so the tone control should work, if there are no dried out caps, although with such low impedances I'm not quite sure how! The unmodified gain of the amplifers is 20, so those 1u feedback caps will behave as if they were about 20u.

Notice that DC supply to the microphones. This makes me think they are electrets

As soon as i can get into the office, I'll snaffle their all-singing cap tester and find out what the actual values are.

The 150n 'dry' caps are a guess based on the amp spec sheet.

The 0u22 ones are actually mis-printed so any dot would be off the edge of the cap. Size-wise they couldn't realistically be 22u, and it wouldn't make sense, bearing in mind the feed caps are only 33u.

All very strange

[zenguitar]

Bravo Folderol.



Andy

[Martin Walker]

And I agree with you that it's a very strange circuit!

Are you sure those 1u caps are not connected between pins 1 and 8 of the LM386?



Martin

[Folderol]

Absolutely. Pins 1, 7, 8 are as naked as a newborn baby.

Also today I did a few active tests in my {cough}kitchen{cough} workshop, using my signal generator (of DIY amp fame) and a modern true RMS multimeter. At these low frequencies they're pretty good.

I worked with an output level of 1V, which is quite reasonable as a headphone level, and as I'm not sure of the correct source impedance, I used 600ohms, 2.2k and 47k. The result was entirely consistent across all of these, which is something I didn't expect actually.

I can confirm the tone control has no measurable effect at all over the range 40Hz to 4kHz

Also, the -3dB point relative to 400Hz is 70Hz, which is way too high for a bass.

I might try tapping caps across those present (particularly those 33u coupling ones) and see what transpires.

I'm not at work tomorrow, but if I'm down that way I might drop in and see if I can borrow the cap tester - I should probably get one myself

[Dr R]

Huge thanks to Folderol for taking this on. It's beyond the limit of any electronics I have ever forgotten. Which is the tone pot pair??

I think we've found someone who can measure the microphone impedances as requested in a PM. Will let the forum know the outcome.

Thanks again to everyone!

Rich

[Folderol]

OK, there is an error in tracing the circuit, and it now makes much more sense


(Updated image -- HR)

If you click on the link itself, you'll get the corrected drawing. The 1u caps are compensation ones that go from the top of the volume control to the wiper.

This doesn't invalidate any of the actual tests. Also the latest news is that the 0.47u caps are dead, tapping one across them immediately made the tone control work.

The 'dry' cap and the 33u one are effectively forming a 2nd order high pass filter, both of them are dropping a significant voltage at 70Hz. Using an R/C calculator, the 33u one against the volume pot at full volume should give a -3dB point of 4.8Hz! so it looks like those are also dead-ish and the only ones that's supposed to be affecting the roll off should be the 'dry' ones. I only guessed their values, but with the behavior shown and good caps elsewhere I would expect the -3dB point to be nearer 45Hz - much more reasonable.

I'm wondering if the unit has been left somewhere hot for an extended period, or is much older than I thought.

[Hugh Robjohns]

Pretty sure they are going to be electret mics given the DC supply present on the mic lines. All electrets have an active impedance converter-cum-output-driver. So they will have a very low output impedance of the order of 100 Ohms or less.

H

[Martin Walker]

Aha - the 1uF caps now make FAR more sense Folderol!

Martin

[Dr R]

The instrument lives in north Wales - but does travel to South Africa and France for extended periods so heat damage is entirely plausible.

As an aside: Apparently it's great to travel with as it's light and will go in an aircraft overhead locker. The downside is that in its bag it looks like a rifle and airport security get very twitchy

[ef37a]

Well done that man!

Are you sure those capacitors are not old Tantalums? They promised much but after just a few years generally delivered shorts! I guess a few also went open circuit.

Dave.

[Folderol]

@hugh
The tone control tests suggest that these mics are round about 1k. Even a 600ohm source means the control has less effect. 100ohm would give a -3dB point of 3.4kHz at maximum attenuation.

@Dave
That was my first thought, but they look like standard SM electrolytics, and the codes also suggest that. I'll be replacing the 0u47 ones with multilayer ceramics. They are not much bigger, more stable and much quieter.

I'm a bit concerned about the volume pot wiring, as it will have a frequency component, depending on it's position (as well as what's on the output). Near minimum the LF cutoff will be about 400Hz as only the 10ohm resistor will be relevant against the 33u cap, but at full volume it will be 4.8Hz. This could account of the apparent poor low end. So maybe it should be run as close to max as possible.

And finally a bit of bad news. My sig genny seems to have developed a fault on it's higher ranges and goes unstable when the gain is turned up and it's only about 45 years old.

I'd write a nasty letter to the manufacturer... only it was me

[ef37a]

" My sig genny seems to have developed a fault on it's higher ranges and goes unstable when the gain is turned up"

Sounds like a duff decoupling CAPACITOR to me mate!

Dave.

[Folderol]

Job done now, and the beastie will soon be on its way home.

Latest drawing is here. Incidentally, I tend to post links to PDFs for drawings as they are much clearer than any bit-image. The changes are all indicated as you can see.

Swapping the tone control caps for ceramic ones was a no-brainer. As well as being more stable, they are actually smaller than the originals!

From actual frequency response behaviour I've downgrade my guess at the original 'dry' caps to only 100nF.

Increasing the 33u caps to 47u and adding the 100ohm resistors, makes the worst-case LF -3dB point about 45Hz. That's with the volume control a sniff above zero. As the volume level increases that drops to to about 30Hz. This has one drawback in that the level will be lower in headphones. However, at the same time the battery life will be slightly better as the amplifiers won't be dumping power into a 10ohm resistor at low volume.

I added an extra resistor, rather than changing the value of the existing one partly because space was getting really tight, but also because it's quick and easy to revert if needed.

[Martin Walker]

Job done now, and the beastie will soon be on its way home.

Bravo Folderol

- looking forward to hearing the end result in situ in the instrument

Martin

[Dr R]

Yes, huge thanks again to Folderol for doing this.

Hope to have the board and bass reunited this weekend, though the football may have an impact on available time

I didn't record the original sound, but will definitely post recordings of the new improved sound.

Rich

[Dr R]

Quick update:

"Real Life" has completely got in the way of reuniting the board and bass.

We've therefore booked a meeting room at work over lunch tomorrow and the plan is to bring in the gear to reconnect it and record it then. Fortunately the directors on our floor appreciate that most of the time we actually do useful and productive work:

http://dilbert.com/strip/1994-06-26

Rich

[Folderol]

'Real Life' eh? Not sure I want one of those - sounds expensive

[Drew Stephenson]

I think I saw one from a distance once, but I wanted no part of it...

[Dr R]

Still not done the soldering. There is a small, but at least non-zero, chance of it happening this weekend.

"Real Life" is on average a "good thing", but the peaks and troughs are unpredictable. Maybe there should be some sort of calibration system which allows for dynamics but didn't leave you totally exhausted at the end of two weeks?
LUFS = Life Units relative to a Fortnights' Sleep

Rich

[Drew Stephenson]

"Real Life" is on average a "good thing", but the peaks and troughs are unpredictable. Maybe there should be some sort of calibration system which allows for dynamics but didn't leave you totally exhausted at the end of two weeks?
LUFS = Life Units relative to a Fortnights' Sleep

Rich

Presumably the scale goes into the negative!

[Dr R]

I was thinking more that if there were too many constant peaks at high amplitude it was unpleasant and grating. In which case negative units would imply something very strange that I can't quite think of an analogy for..

[Drew Stephenson]

Dead on your feet.